2D-Haar声学特征超向量快速生成方法

针对大数据量音频的高速处理,提出一种快速的声学特征超向量生成方法,有效提高音频识别系统的识别速度和精度.所提方法首先将多个连续音频帧的常用声学特征构成声学特征图,进而使用低复杂度的运算方法在其中快速提取维数达数十万的Haar-like声学特征;然后使用AdaBoost.MH算法,筛选出具有较高代表性的Haar-like声学特征模式组合,用以构成声学特征超向量;进而提出Random AdaBoost特征筛选方法,进一步提高特征筛选速度.实验结果表明,在音频事件识别、说话人识别、说话人性别识别3种场合下,使用Haar-like声学特征可以使SVM、C5.0、AdaBoost等识别算法获得比MFCC、PLP、LPCC等常用声学特征更高的识别准确率,同时可以获得7~20倍的训练速度提升和5~10倍的识别速度提升.      

基于稀疏表示和粒子滤波的在线目标跟踪算法

针对目标跟踪过程中由于外形变化或者遮挡所造成的跟踪效果下降或导致漂移的问题,提出一种粒子滤波框架下基于稀疏表示的在线目标跟踪算法.采用分层梯度方向直方图(PHOG)特征对目标模板进行描述,并且每一个候选模板都可以通过PHOG基向量和琐碎模板进行稀疏表示,进而利用L1范数最小化方法进行最优求解.为保证在遮挡的情况下目标跟踪的精度,对目标遮挡部分和非遮挡部分进行拆分建模,并利用PCA子空间增量学习的方式不断更新目标跟踪模型.通过对具有挑战性的跟踪视频进行定性和定量分析,实验证明该方法在跟踪精度上要优于传统的跟踪方法.      

一种重复二分CSP4类运动想象脑电信号特征提取算法

针对脑机接口（BCI）系统中4类运动想象的脑电信号ERD/ERS现象进行研究,提出了一种重复二分共同空间模式（RB-CSP）算法用于4类运动想象脑电信号的特征提取,并运用SVM进行分类研究. 实验结果表明,该方法与传统的4分类CSP扩展算法OVR-CSP相比,减小了算法复杂程度,缩短了信号处理时间,提高了准确率,为在线脑机应用提供了一种新的解决方法.      

基于半圆形局部二值模式结构相关性描述子的图像检索

针对传统的结构基元方法缺少对不同量化颜色层中的中心像素点与其潜在邻居的相似性信息描述问题,提出了一种半圆形局部二值模式结构相关性描述子,并将其应用在图像检索中.首先,定义了一种新的半圆形局部二值模式结构基元;其次,检测不同量化颜色层中的结构基元;最后,提取新结构基元的空间分布和对比度特征.相比传统的结构基元方法,提出的描述子检测的结构基元更加丰富,包含更多可能的结构区分性.在不同图像库上的实验结果表明了所提方法的有效性.     

中文文本分类技术比较研究

文本分类中特征选择、权重计算及分类算法三个阶段中都存在一些经典方法,在实际的中文文本分类任务中,如何从各阶段不同方法的组合中找到一个好的组合成为值得研究的问题。比较研究中文文本分类中各阶段经典方法的不同组合对分类效果的影响结果表明:采用CHI特征选择方法、TFIDF权重计算方法及SVM分类方法的组合为最佳组合。     

An Analysis of a Passage from the Stylistic Perspective

The thesis analyzes Churchill＇s Speech on Hitler＇s Invasion upon the USSR from the perspective of stylistics. Through the analysis of grammatical features, lexical features, lexical features, cohesion and context, and figures of speech, the author intends to view the article from a more scientific approach, and at the same time, reveal the important role of linguistic approach in literary evaluation.     

基于视觉注意力模型的FCM法在医学诊断中的应用

基于人类视觉注意模型得到局部视觉显著图,然后以注意焦点初始化聚类中心进行模糊C均值聚类,从而实现图像ROI的提取.实验证明基于图像视觉注意力模型的FCM方法应用于计算机辅助诊断中可以提高医学诊断的有效性和快速性.     

Related‐variables selection in temporal disaggregation

Two related‐variables selection methods for temporal disaggregation are proposed. In the first method, the hypothesis tests for a common feature (cointegration or serial correlation) are first performed. If there is a common feature between observed aggregated series and related variables, the conventional Chow–Lin procedure is applied. In the second method, alternative Chow–Lin disaggregating models with and without related variables are first estimated and the corresponding values of the Bayesian information criterion (BIC) are stored. It is determined on the basis of the selected model whether related variables should be included in the Chow–Lin model. The efficacy of these methods is examined via simulations and empirical applications. Copyright © 2008 John Wiley & Sons, Ltd.     

High speed robust image registration and localization using optimized algorithm and its performances evaluation

Local invariant algorithm applied in downward-looking image registration, usually computes the camera’s pose relative to visual landmarks. Generally, there are three requirements in the process of image registration when using these approaches. First, the algorithm is apt to be influenced by illumination. Second, algorithm should have less computational complexity. Third, the depth information of images needs to be estimated without other sensors. This paper investigates a famous local invariant feature named speeded up robust feature (SURF), and proposes a highspeed and robust image registration and localization algorithm based on it. With supports from feature tracking and pose estimation methods, the proposed algorithm can compute camera poses under different conditions of scale, viewpoint and rotation so as to precisely localize object’s position. At last, the study makes registration experiment by scale invariant feature transform (SIFT), SURF and the proposed algorithm, and designs a method to evaluate their performances. Furthermore, this study makes object retrieval test on remote sensing video. For there is big deformation on remote sensing frames, the registration algorithm absorbs the Kanade-Lucas-Tomasi (KLT) 3-D coplanar calibration feature tracker methods, which can localize interesting targets precisely and efficiently. The experimental results prove that the proposed method has a higher localization speed and lower localization error rate than traditional visual simultaneous localization and mapping (vSLAM) in a period of time.     

英语语调的副语言特征与话语理解

人类交际过程是一个复杂的过程。英语语调是一种副语言形式，不同的语调传递不同的信息和话语意义，表达不同的态度或情感，它具有特殊的交际功能与价值。本文主要探讨英语语调的副语言特征，帮助学生正确理解语调的话语功能，推断话语的真正含义。